High Performance. Energy Efficiency. Low TCO.
The AMD EPYC 9004 Server CPU can help power enterprise workloads for high performance, energy efficiency, and seamless x86 compatibility across on-premises and cloud-native deployments. Minimize your total cost of ownership (TCO) while maximizing results across a wide range of workloads.
“Zen 4” Architecture
Built with up to 128 “Zen 4” or “Zen 4c” cores and 12 DDR5 memory channels for exceptional memory bandwidth and capacity. The 4th Gen AMD EPYC CPU architecture enables high performance, energy-efficient solutions optimized for a wide range of data centers.
~65%
More Total AI Use Case Throughput/min
*TPCx-AI SF30 2P 96C EPYC™ 9654 vs 2P 64C Xeon 8582+-¹
Enhance AI Workloads
Deliver leadership performance and efficiency for demanding AI workloads with platforms powered by AMD EPYC 9004 Server CPUs—helping enterprises and cloud providers scale AI infrastructure seamlessly into the future.
Cloud Computing Solutions
Deliver leadership performance and efficiency for demanding workloads with Cloud Virtual Machines (VMs) powered by AMD EPYC 9004 Server CPUs - helping cloud providers scale easily into the future.
Use
~38%
Fewer Instances
to process 7 million requests per second running NGINX™ on AMD EPYC CPU powered Amazon EC2 M7a.4xl vs. Intel Xeon powered M7i.4xl.2
Up to
~60%
increase enterprise critical operation throughput – maintain SLAs³
*SPECjbb®2015-MultiJVM Critical-jOPS 2P 96C EPYC™ 9684X vs. 2P 64C Xeon Platinum 8592+
Enterprise
Advance your general-purpose servers to achieve exceptional value across power efficiency, total cost of ownership (TCO), space savings, and performance. AMD EPYC 9004 Server CPU is designed for all enterprise customers.
Partner Solutions
AMD works with a variety of partners to design and develop optimized solutions featuring 4th Gen AMD EPYC™.
Model Specifications
- 9004 Series Processors
- 9004 Series with 3D V-Cache™ Technology
- 97x4 Processors
Support & Resources
Footnotes
- SP5-051A: TPCx-AI SF30 derivative workload comparison based on AMD internal testing running multiple VM instances as of 4/13/2024. The aggregate end-to-end AI throughput test is derived from the TPCx-AI benchmark and as such is not comparable to published TPCx-AI results, as the end-to-end AI throughput test results do not comply with the TPCx-AI Specification. AMD system configuration: Processors: 2 x AMD EPYC 9654; Frequencies: 2.4 GHz | 3.7 GHz; Cores: 96 cores per socket (1 NUMA domain per socket); L3 Cache: 384MB/socket (768MB total); Memory: 1.5TB (24) Dual-Rank DDR5-5600 64GB DIMMs, 1DPC (Platform supports up to 4800MHz); NIC: 2 x 100 GbE Mellanox CX-5 (MT28800); Storage: 3.2 TB Samsung MO003200KYDNC U.3 NVMe; BIOS: 1.56; BIOS Settings: SMT=ON, Determinism=Power, NPS=1, PPL=400W, Turbo Boost=Enabled; OS: Ubuntu® 22.04.3 LTS; Test config: 6 instances, 64 vCPUs/instance, 2663 aggregate AI use cases/min vs. Intel system configuration: Processors: 2 x Intel® Xeon® Platinum 8592+; Frequencies: 1.9 GHz | 3.9 GHz; Cores: 64 cores per socket (1 NUMA domain per socket); L3 Cache: 320MB/socket (640MB total); Memory: 1TB (16) Dual-Rank DDR5-5600 64GB DIMMs, 1DPC; NIC: 4 x 1GbE Broadcom NetXtreme BCM5719 Gigabit Ethernet PCIe; Storage: 3.84TB KIOXIA KCMYXRUG3T84 NVMe; BIOS: ESE124B-3.11; BIOS Settings: Hyperthreading=Enabled, Turbo boost=Enabled, SNC=Disabled; OS: Ubuntu® 22.04.3 LTS; Test config: 4 instances, 64 vCPUs/instance, 1607 aggregate AI use cases/min. Results may vary due to factors including system configurations, software versions and BIOS settings. TPC, TPC Benchmark and TPC-C are trademarks of the Transaction Processing Performance Council.
- SP5C-013: Based on AMD testing as of 9/28/2023, M7a.4xl has 1.62x performance advantage compared to M7i.4xl for NGINX WRK testing throughput. Instance pricing of m7i.4xlarge: $0.8064 / hr and m7a.4xlarge: $0.92736 / hr is based on published Amazon EC2 On-Demand Pricing for US East (Ohio) Region as of 9/28/2023. https://calculator.aws/#/createCalculator/ec2-enhancement
Performance (NGINX WRK requests/sec (RPS) )
m7i.4xlarge 893621.554
m7a.4xlarge 1448376.598
Cloud performance results presented are based on the test date in the configuration. Results may vary due to changes to the underlying configuration, and other conditions such as the placement of the VM and its resources, optimizations by the cloud service provider, accessed cloud regions, co-tenants, and the types of other workloads exercised at the same time on the system. NGINX is a trademark of Nginx Software Inc.
- SP5-104B: SPECjbb® 2015-MultiJVM Critical-jOPS based on published scores from www.spec.org as of 1/4/2024. Configurations: 2P AMD EPYC 9684X (679,119 SPECjbb®2015 MultiJVM critical-jOPS, 783,622 SPECjbb®2015 MultiJVM max-jOPS, 192 Total Cores, https://spec.org/jbb2015/results/res2023q3/jbb2015-20230906-01166.html ) is 1.61x the critical-jOPS performance published 2P 64C Intel Xeon Platinum 8592+ (421,207 SPECjbb®2015 MultiJVM critical-jOPS, 480,007 SPECjbb®2015 MultiJVM max-jOPS, 128 Total Cores, https://spec.org/jbb2015/results/res2024q1/jbb2015-20231224-01210.html ). SPEC® and SPECjbb® are registered trademarks of the Standard Performance Evaluation Corporation. See www.spec.org for more information.
Footnotes
- SP5-051A: TPCx-AI SF30 derivative workload comparison based on AMD internal testing running multiple VM instances as of 4/13/2024. The aggregate end-to-end AI throughput test is derived from the TPCx-AI benchmark and as such is not comparable to published TPCx-AI results, as the end-to-end AI throughput test results do not comply with the TPCx-AI Specification. AMD system configuration: Processors: 2 x AMD EPYC 9654; Frequencies: 2.4 GHz | 3.7 GHz; Cores: 96 cores per socket (1 NUMA domain per socket); L3 Cache: 384MB/socket (768MB total); Memory: 1.5TB (24) Dual-Rank DDR5-5600 64GB DIMMs, 1DPC (Platform supports up to 4800MHz); NIC: 2 x 100 GbE Mellanox CX-5 (MT28800); Storage: 3.2 TB Samsung MO003200KYDNC U.3 NVMe; BIOS: 1.56; BIOS Settings: SMT=ON, Determinism=Power, NPS=1, PPL=400W, Turbo Boost=Enabled; OS: Ubuntu® 22.04.3 LTS; Test config: 6 instances, 64 vCPUs/instance, 2663 aggregate AI use cases/min vs. Intel system configuration: Processors: 2 x Intel® Xeon® Platinum 8592+; Frequencies: 1.9 GHz | 3.9 GHz; Cores: 64 cores per socket (1 NUMA domain per socket); L3 Cache: 320MB/socket (640MB total); Memory: 1TB (16) Dual-Rank DDR5-5600 64GB DIMMs, 1DPC; NIC: 4 x 1GbE Broadcom NetXtreme BCM5719 Gigabit Ethernet PCIe; Storage: 3.84TB KIOXIA KCMYXRUG3T84 NVMe; BIOS: ESE124B-3.11; BIOS Settings: Hyperthreading=Enabled, Turbo boost=Enabled, SNC=Disabled; OS: Ubuntu® 22.04.3 LTS; Test config: 4 instances, 64 vCPUs/instance, 1607 aggregate AI use cases/min. Results may vary due to factors including system configurations, software versions and BIOS settings. TPC, TPC Benchmark and TPC-C are trademarks of the Transaction Processing Performance Council.
- SP5C-013: Based on AMD testing as of 9/28/2023, M7a.4xl has 1.62x performance advantage compared to M7i.4xl for NGINX WRK testing throughput. Instance pricing of m7i.4xlarge: $0.8064 / hr and m7a.4xlarge: $0.92736 / hr is based on published Amazon EC2 On-Demand Pricing for US East (Ohio) Region as of 9/28/2023. https://calculator.aws/#/createCalculator/ec2-enhancement
Performance (NGINX WRK requests/sec (RPS) )
m7i.4xlarge 893621.554
m7a.4xlarge 1448376.598
Cloud performance results presented are based on the test date in the configuration. Results may vary due to changes to the underlying configuration, and other conditions such as the placement of the VM and its resources, optimizations by the cloud service provider, accessed cloud regions, co-tenants, and the types of other workloads exercised at the same time on the system. NGINX is a trademark of Nginx Software Inc. - SP5-104B: SPECjbb® 2015-MultiJVM Critical-jOPS based on published scores from www.spec.org as of 1/4/2024. Configurations: 2P AMD EPYC 9684X (679,119 SPECjbb®2015 MultiJVM critical-jOPS, 783,622 SPECjbb®2015 MultiJVM max-jOPS, 192 Total Cores, https://spec.org/jbb2015/results/res2023q3/jbb2015-20230906-01166.html ) is 1.61x the critical-jOPS performance published 2P 64C Intel Xeon Platinum 8592+ (421,207 SPECjbb®2015 MultiJVM critical-jOPS, 480,007 SPECjbb®2015 MultiJVM max-jOPS, 128 Total Cores, https://spec.org/jbb2015/results/res2024q1/jbb2015-20231224-01210.html ). SPEC® and SPECjbb® are registered trademarks of the Standard Performance Evaluation Corporation. See www.spec.org for more information.